Focus on the Core: Empowering Diffusion Large Language Models by Self-Contrast

• URL: http://arxiv.org/abs/2605.01373v1
• Date: Sat, 02 May 2026 10:46:56 GMT
• Title: Focus on the Core: Empowering Diffusion Large Language Models by Self-Contrast
• Abstract summary: We investigate high-information-density (HD) tokens and present two key findings.<n>We propose Focus on the Core textbf(FoCore), a training-free decoding strategy.<n>FoCore consistently improves generation quality and efficiency across both LLaDA and Dream backbones.
• Score: 25.00502163665476
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The iterative denoising paradigm of Diffusion Large Language Models (DLMs) endows them with a distinct advantage in global context modeling. However, current decoding strategies fail to leverage this capability, typically exhibiting a local preference that overlooks the heterogeneous information density within the context, ultimately degrading generation quality. To address this limitation, we systematically investigate high-information-density (HD) tokens and present two key findings: (1) explicitly conditioning on HD tokens substantially improves output quality; and (2) HD tokens exhibit an early-decoding tendency, converging earlier than surrounding tokens. Motivated by these findings, we propose Focus on the Core \textbf{(FoCore)}, a training-free decoding strategy that utilizes HD tokens in a self-contrast manner, wherein HD tokens are temporarily remasked as negative samples, to guide generation. We further introduce FoCore\_Accelerate \textbf{(FoCore\_A)}, an efficient variant that, upon detecting HD token convergence, performs parallel decoding over stable candidates within a local context window, substantially accelerating generation. Extensive experiments on math, code and logical reasoning benchmarks demonstrate that FoCore consistently improves generation quality and efficiency across both LLaDA and Dream backbones. For instance, on HumanEval, FoCore improves pass@1 from 39.02 to 42.68 over standard Classifier-Free Guidance, while FoCore-A reduces the number of decoding steps by 2.07x and per-sample latency from 20.76s to 8.64s (-58.4\%).

Related papers

• SimSD: Simple Speculative Decoding in Diffusion Language Models [61.33773959352141]
  Diffusion large language models (dLLMs) offer faster inference through parallel or blockwise decoding.<n>Their masked language modeling formulation remains incompatible with standard token-level speculative decoding.<n>We propose a speculative decoding algorithm for dLLMs, named SimSD, that equips dLLMs with temporally valid token-level contexts.<n>Our method achieves up to 7.46x higher decoding throughput while maintaining and even improving average generation quality.
  arXiv  Detail & Related papers  (2026-06-01T17:46:46Z)
• Breaking Block Boundaries: Anchor-based History-stable Decoding for Diffusion Large Language Models [54.468024377748]
  We propose Anchor-based History-stable Decoding (AHD) as a training-free, plug-and-play dynamic decoding strategy.<n>AHD monitors the stability trend of tokens in real time through dynamic anchors.<n>Our approach reduces decoding steps by 80% while improving performance by 3.67%.
  arXiv  Detail & Related papers  (2026-04-10T05:08:39Z)
• CaTok: Taming Mean Flows for One-Dimensional Causal Image Tokenization [122.88484422855934]
  We present CaTok, a 1D causal image tokenizer with a MeanFlow decoder.<n>By selecting tokens over time intervals, CaTok learns causal 1D representations that support both fast one-step generation and high-fidelity multi-step sampling.<n> Experiments demonstrate that CaTok achieves state-of-the-art results on ImageNet reconstruction, reaching 0.75 FID, 22.53 PSNR and 0.674 SSIM.
  arXiv  Detail & Related papers  (2026-03-06T16:39:17Z)
• CORE: Context-Robust Remasking for Diffusion Language Models [51.59514489363897]
  We propose Context-Robust Remasking (CORE), a training-free framework for inference-time revision.<n>Rather than trusting static token probabilities, CORE identifies context-brittle tokens by probing their sensitivity to targeted masked-context perturbations.<n>On LLaDA-8B-Base, CORE delivers consistent improvements across reasoning and code benchmarks, outperforming compute-matched baselines and improving MBPP by up to 9.2 percentage points.
  arXiv  Detail & Related papers  (2026-02-04T00:12:30Z)
• Deferred Commitment Decoding for Diffusion Language Models with Confidence-Aware Sliding Windows [33.361153168706444]
  We propose Deferred Commitment Decoding (DCD) as a training-free decoding strategy.<n>DCD maintains a confidence-aware sliding window over masked tokens, resolving low-uncertainty tokens early while deferring high-uncertainty tokens until sufficient contextual evidence becomes available.<n>Experiments show that DCD improves generation accuracy by 1.39% with comparable time on average compared to fixed block-based diffusion methods, with the most significant improvement reaching 9.0%.
  arXiv  Detail & Related papers  (2026-01-05T12:57:33Z)
• Accelerating Diffusion LLM Inference via Local Determinism Propagation [27.751279909685604]
  LocalLeap is a training-free adaptive parallel decoding strategy.<n>It achieves 6.94$times$ throughput improvements and reduces decoding steps to just 14.2% of the original requirement.
  arXiv  Detail & Related papers  (2025-10-08T14:39:34Z)
• Fast and Fluent Diffusion Language Models via Convolutional Decoding and Rejective Fine-tuning [23.58934174168992]
  Autoregressive (AR) language models generate text one token at a time, which limits their inference speed.<n>We propose Convolutional decoding (Conv), a normalization-based method that narrows the decoding window without hard segmentation.<n>We also introduce Rejecting Rule-based Fine-Tuning (R2FT), a post-hoc training scheme that better aligns tokens at positions far from context.
  arXiv  Detail & Related papers  (2025-09-18T17:48:21Z)
• LayerCake: Token-Aware Contrastive Decoding within Large Language Model Layers [53.43862310647276]
  Large language models (LLMs) excel at natural language understanding and generation but remain vulnerable to factual errors.<n>We introduce a token-aware, layer-localized contrastive decoding method that aligns specific token types with their most influential transformer layers to improve factual generation.<n>Our method requires no additional training or model modification, and experiments demonstrate that our method consistently improves factuality across multiple LLMs and various benchmarks.
  arXiv  Detail & Related papers  (2025-07-06T14:35:43Z)
• Token-Shuffle: Towards High-Resolution Image Generation with Autoregressive Models [92.18057318458528]
  Token-Shuffle is a novel method that reduces the number of image tokens in Transformer.<n>Our strategy requires no additional pretrained text-encoder and enables MLLMs to support extremely high-resolution image synthesis.<n>In GenAI-benchmark, our 2.7B model achieves 0.77 overall score on hard prompts, outperforming AR models LlamaGen by 0.18 and diffusion models LDM by 0.15.
  arXiv  Detail & Related papers  (2025-04-24T17:59:56Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.